The etiology of germ cell tumors (GCT) is poorly understood. Previous studies, mainly focused on testicular cancer and conducted among adult populations, have suggested that certain pre-natal exposures, such a maternal exogenous hormone use and high endogenous hormone level during pregnancy, and parental occupational exposures, may be associated with an increased risk of GCT. These adult studies, however, were limited because: a) the time between pre-natal exposures and GCT diagnosis is long and validation of pre-natal exposure was not feasible; b) post-natal exposures, particularly hormone-related factors and occupational exposures were usually not adjusted; c) only one type of GCT was studied in virtually all previous studies, which precluded a comparison of the risk factors for different types of GCT. This proposed study is designed to investigate specific and shared risk factors for three types of GCT (testicular, ovarian, and non-gonadal) and to test the following hypotheses: (1) Maternal exogenous hormone use and high endogenous estrogen level are associated with an increase risk of GCT; (2) Parental occupational exposures are related to the risk of GCT in offspring; (3) Expression of human long repetitive elements (L1Hs) defines a sub-group of GCT cases with specific etiologic, histologic and prognostic features; (4) Male GCT cases have a high frequency of constitutional chromosome abnormalities, and risk factors for GCT are different between GCTs with and without constitutional chromosome abnormalities. Utilizing the unique resources available through the Children's Cancer Group, this study aims to (1) recruit 615 childhood GCT cases and 836 frequency matched controls, (2) interview parents of all study subjects using a comprehensive questionnaire eliciting information on exposures of index child in pre-conception, pre-natal and post-natal periods, and validate maternal exposures during 6 months prior to and during index pregnancy through review of medical records, (3) collect tumor tissue of GCT cases for measuring L1Hs expression and tumor karyotype, and (4) examine the constitutional chromosome abnormalities for all GCT cases. This study represents the first large epidemiologic investigation to study risk factors of childhood GCT and to examine the interaction between genetic and environmental factors in the development of GCT. The relatively short interval between exposures and GCT diagnosis will facilitate recall of pre-natal exposures and allow for validation of exposures. The relatively large sample size will provide sufficient statistical power to address the hypotheses being investigated.